National Oceanic and
Atmospheric Administration
United States Department of Commerce


 

FY 1990

Beaufort Sea mesoscale circulation study—Final Report

Aagaard, K., C.H. Pease, A.T. Roach, and S.A. Salo

NOAA Tech. Memo. ERL PMEL-90, NTIS: PB90–158775, 114 pp (1989)


The Beaufort Sea Mesoscale Project was undertaken to provide a quantitative understanding of the circulation over the Beaufort Sea shelf and of its atmospheric and oceanic forcing. Major emphasis has been placed on providing extensive synoptic oceanographic and meteorological coverage of the Alaskan Beaufort Sea during 1986-88. In addition, supplementary measurements have been made in the southern upstream waters of Bering Strait and the Chukchi Sea. The work has resulted in an unprecedented regional data set for both the ocean and the atmosphere. The principal conclusions are as follows: 1) Below the upper 40-50 m of the ocean, the major circulation feature of the outer shelf and slope is the Beaufort Undercurrent, a strong flow which is directed eastward in the mean, but which is subject to frequent reversals toward the west. The reversals are normally associated with upwelling onto the outer shelf. The undercurrent is very likely part of a basin-scale circulation within the Arctic Ocean. 2) While we find statistically significant wind influence on the subsurface flow in the southern Beaufort Sea, it is generally of secondary importance, accounting for less than 25% of the flow variance below 60 m. An important implication is that at least below the mixed layer, the circulation on the relatively narrow Beaufort shelf is primarily forced by the ocean rather than by the local wind. This oceanic forcing includes shelf waves and eddies. Therefore, to the extent that a localized problem or process study requires consideration of the shelf circulation, such as would be the case for oil-spill trajectory modeling, a larger-scale framework must be provided, within which the more local problem may be nested. 3) There were large changes in wind variance with season, with the largest variances occurring in the late summer/early autumn and again in January because of blocking ridges in the North Pacific shifting the storm track westward over the west coast of Alaska and across the North Slope. 4) Despite the seasonally varying wind field, as well as the large seasonal differences in the upper-ocean temperature and salinity fields, we find no evidence for a seasonal variability in the subsurface circulation in the Beaufort Sea. This situation contrasts with that in Bering Strait and probably in the Chukchi Sea, where a seasonal cycle in the transport is apparent. Therefore, while the northward flow of water from the Pacific is of major significance to the structure and chemistry of the upper ocean in the Arctic (including the Beaufort Sea), as well as its ice cover and biota, the dynamic significance of that flow to the Beaufort Sea appears small. 5) In contrast to the lack of a seasonal oceanographic signal at depth, the interannual variability in the flow characteristics can be considerable. For example, during the period fall 1986-spring 1987, the Beaufort Undercurrent appears to have been deeper by 30-40 m compared with both earlier and ensuing measurements. The consequences of such anomalies for the upper-ocean velocity structure and transport are likely significant. 6) During much of the experiment, the meteorological conditions were milder than normal, consistent with less coastal ice in the summer and autumn, the passage of more storms up the west coast of Alaska and across the North Slope, and generally higher air temperatures along the North Slope. These climatological near-minimum ice years were followed in 1988 by the heaviest summer ice along the Chukchi coast since 1975. 7) The atmospheric sea-level pressure field was well represented by the METLIB products from the FNOC surface analysis if the 12-hour lag of the FNOC pressures was taken into account. However, the FNOC surface air temperature field does not accurately represent either the land-based stations or the drifting ice buoys. The errors in the FNOC temperature field showed a systematic over-prediction during winter and spring of 10-20°C, leading to an annual over-prediction of air temperature by 3–13°C at all sites. Gradient winds from FNOC are therefore well suited for modeling purposes if they are calculated from the time-shifted surface analysis, but the FNOC surface temperature analysis should not be used for any model calculations, except perhaps as an upper boundary condition for a rather complete planetary boundary layer model.




Feature Publications | Outstanding Scientific Publications

Contact Sandra Bigley |